What is the Edge?
Despite much discussion about Edge, the concept should really be considered a deployment strategy rather than a new technology as some are inclined to suggest. As a concept Edge is not new and has been around for multiple decides. Consider Akamai’s Content Distribution Network (CDN), which was being built in the late 1990’s and is a clear example of an Edge deployment in all but name.
In reality ‘Edge’ is something of a marketing term used to describe a strategy using technology that is not necessarily new. It is also evident that ‘Edge’ means very different things to different people and organisations, depending upon their business requirements and technology base. Equipment deployments ranging from 20 Watts (a so-called Nano Data Centre), to over 5 MW have been described as ‘Edge’ sites.
It could be argued that the term “Edge Data Centre” does not really relate to a new phenomenon, but is rather a collective name for several different types of data centres which serve a number of different purposes, none of which are new, rather that these have been relatively unrecognised or “unclassified”, to date. These are likely to be relatively small remote sites which might include responses to increasing data sovereignty requirements, expectations for reductions in latency and proximal workload processing.
Others would argue that the ‘Edge’ is entirely new and relates to very specific and separate elements within the ‘Cloud’, which offer new functions and features, particularly in relation to IoT and mobile services. Either way there is a clear perception that Edge sites are different to the more traditional ‘Core’ data centre and therefore need to be maintained and managed in a different way. This will vary depending on the technology deployed and an organisation’s business objectives and requirements.
A better and more explanatory term for Edge might be ‘Proximal Processing and Storage”, with the intention being to place services closer to the end user for a variety of reasons which might include one of more of the following: application and network latency, network transmission costs, data sovereignty requirements, data protection requirements, advantageous taxation regimes and more widespread data gathering.